Enter opens the highlighted result. Esc closes.

Qwen2.5 72B Instruct vs Qwen3 32B

Qwen2.5 72B Instruct leads Qwen3 32B on the Artificial Analysis Intelligence Index, 7.7 to 7.3, at about the same output price ($0.40 vs $0.40 per million). Updated October 22, 2026; 38 benchmarks have a result for both.

Replace Qwen2.5 72B Instruct Replace Qwen3 32B
Qwen2.5 72B InstructQwen
Coding
No data
Overall
26.5
Output $/M
$0.40
4
Qwen3 32BQwen
Coding
No data
Overall
40.8
Output $/M
$0.40
34
vs boards won of 38 shared

Qwen2.5 72B Instruct or Qwen3 32B: which to pick

Neither model has a coding score yet, so the two cannot be ranked on coding.

Pick Qwen2.5 72B Instruct for work measured by LiveCodeBench (Artificial Analysis run) or LMArena Text: coding, where it leads.

Model AQwen2.5 72B InstructQwenNot benchmarkedNo coding benchmarkNo cost per solved taskModel BQwen3 32BQwenNot benchmarkedNo coding benchmarkNo cost per solved task
Shared results
38
Qwen2.5 72B Instruct only
15
Qwen3 32B only
7
Like-for-like categories
7 of 9

Only benchmarks that tested both models count toward a pick; a result only one model has is shown but not compared.

Which one for your work

Each card is decided by one number both models carry. Where only one has it, the card says so and makes no call.

At a glance

Coding index
Qwen2.5 72B Instruct not benchmarked, Qwen3 32B not benchmarked.
Widest category
Math, 43.6 points apart on average over 5 shared benchmarks.
Widest single gap
AA Math Index: Qwen2.5 72B Instruct 14.0 against Qwen3 32B 73.0, 59.0 points apart on a 0 to 100 scale.
Output price
Both cost $0.40 per million output tokens.
Context window
Qwen2.5 72B Instruct takes 33K tokens, Qwen3 32B 131K.

Benchmark by benchmark

First the score: boards each model leads, per category, where both have a result. Then every board, a category to a fold; bars are each result on a 0 to 100 scale, so compare the two in a row, not one row with another.

Boards won, where both have a result
CategoryQwen2.5 72B InstructBarsQwen3 32BTies
Coding13
Intelligence11
Agentic work10
Reasoning06
Math05
Knowledge05
General114
All shared boards434
CodingLike-for-likeSix of these boards feed the Vibe Coding Index13
Coding benchmarks for Qwen2.5 72B Instruct and Qwen3 32B
BenchmarkQwen2.5 72B InstructQwen3 32BLeader
Aider polyglotNot measured40.0%Only Qwen3 32B measured
LiveCodeBench (Artificial Analysis run)27.6%54.6%Qwen3 32B
SciCodeNot measured36.0%Only Qwen3 32B measured
Terminal-Bench Hard (Artificial Analysis run)4.5%3.0%Qwen2.5 72B Instruct
WeirdML16.0%Not measuredOnly Qwen2.5 72B Instruct measured
LMArena Text: coding12921357Qwen3 32B
LMArena Text: software and IT12921366Qwen3 32B
Neither model has a result on the other 32 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index11
Intelligence benchmarks for Qwen2.5 72B Instruct and Qwen3 32B
BenchmarkQwen2.5 72B InstructQwen3 32BLeader
AA Intelligence Index7.77.3Qwen2.5 72B Instruct
Epoch Capabilities Index129.0138.4Qwen3 32B
Agentic workLike-for-likePublic boards, shown as published, kept out of the index10
Agentic work benchmarks for Qwen2.5 72B Instruct and Qwen3 32B
BenchmarkQwen2.5 72B InstructQwen3 32BLeader
τ²-Bench Telecom34.5%29.8%Qwen2.5 72B Instruct
METR time horizons35.8%Not measuredOnly Qwen2.5 72B Instruct measured
BALROG16.2%Not measuredOnly Qwen2.5 72B Instruct measured
TheAgentCompany5.7%Not measuredOnly Qwen2.5 72B Instruct measured
Berkeley Function Calling Leaderboard (BFCL)Not measured48.7%Only Qwen3 32B measured
BrowseComp-PlusNot measured10.7%Only Qwen3 32B measured
Neither model has a result on the other 35 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index06
Reasoning benchmarks for Qwen2.5 72B Instruct and Qwen3 32B
BenchmarkQwen2.5 72B InstructQwen3 32BLeader
GPQA Diamond (Artificial Analysis run)49.1%66.8%Qwen3 32B
Humanity's Last Exam (Artificial Analysis run)3.6%7.4%Qwen3 32B
GPQA Diamond (Epoch AI run)49.1%65.7%Qwen3 32B
BIG-Bench Hard79.8%Not measuredOnly Qwen2.5 72B Instruct measured
LMArena Text: hard prompts12711334Qwen3 32B
LMArena Text: expert12451362Qwen3 32B
LMArena Text: hard prompts in English12791350Qwen3 32B
Kagi LLM BenchmarkNot measured48.7%Only Qwen3 32B measured
Neither model has a result on the other 14 boards here
MathLike-for-likePublic boards, shown as published, kept out of the index05
Math benchmarks for Qwen2.5 72B Instruct and Qwen3 32B
BenchmarkQwen2.5 72B InstructQwen3 32BLeader
AA Math Index14.073.0Qwen3 32B
AIME 2025 (Artificial Analysis run)14.0%73.0%Qwen3 32B
OTIS Mock AIME8.1%66.9%Qwen3 32B
MATH level 563.2%Not measuredOnly Qwen2.5 72B Instruct measured
LMArena Text: math12831399Qwen3 32B
LMArena Text: mathematical work12901420Qwen3 32B
Neither model has a result on the other 7 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index05
Knowledge benchmarks for Qwen2.5 72B Instruct and Qwen3 32B
BenchmarkQwen2.5 72B InstructQwen3 32BLeader
MMLU Pro (Artificial Analysis run)72.0%79.8%Qwen3 32B
MMLU85.3%Not measuredOnly Qwen2.5 72B Instruct measured
AI2 Reasoning Challenge94.5%Not measuredOnly Qwen2.5 72B Instruct measured
LMArena Text: science12751372Qwen3 32B
LMArena Text: medicine and healthcare12581368Qwen3 32B
LMArena Text: legal and government12881324Qwen3 32B
LMArena Text: business and finance12531318Qwen3 32B
Neither model has a result on the other 11 boards here
Web designNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Web design benchmarks for Qwen2.5 72B Instruct and Qwen3 32B
BenchmarkQwen2.5 72B InstructQwen3 32BLeader
Neither model has a result on 9 boards here
GeneralLike-for-likePublic boards, shown as published, kept out of the index114
General benchmarks for Qwen2.5 72B Instruct and Qwen3 32B
BenchmarkQwen2.5 72B InstructQwen3 32BLeader
IFBench36.9%36.3%Qwen2.5 72B Instruct
Long context reasoning (AA-LCR)Not measured0.0%Only Qwen3 32B measured
DTBench62.9%67.5%Qwen3 32B
LMCA13.4%17.3%Qwen3 32B
Fiction.LiveBenchNot measured74.2%Only Qwen3 32B measured
HellaSwag84.8%Not measuredOnly Qwen2.5 72B Instruct measured
WinoGrande82.3%Not measuredOnly Qwen2.5 72B Instruct measured
PIQA82.6%Not measuredOnly Qwen2.5 72B Instruct measured
LMArena Text12691340Qwen3 32B
LMArena Text: instruction following12541305Qwen3 32B
LMArena Text: creative writing12211297Qwen3 32B
LMArena Text: English12831359Qwen3 32B
LMArena Text: non-English12521317Qwen3 32B
LMArena Text: Chinese12731357Qwen3 32B
LMArena Text: German12341341Qwen3 32B
LMArena Text: Japanese1180Not measuredOnly Qwen2.5 72B Instruct measured
LMArena Text: Korean1188Not measuredOnly Qwen2.5 72B Instruct measured
LMArena Text: French1280Not measuredOnly Qwen2.5 72B Instruct measured
LMArena Text: Spanish1255Not measuredOnly Qwen2.5 72B Instruct measured
LMArena Text: Russian12641311Qwen3 32B
LMArena Text: multi-turn12721331Qwen3 32B
LMArena Text: longer queries12811327Qwen3 32B
LMArena Text: writing and language12471298Qwen3 32B
LMArena Text: entertainment and media12111279Qwen3 32B
Neither model has a result on the other 22 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for Qwen2.5 72B Instruct and Qwen3 32B
BenchmarkQwen2.5 72B InstructQwen3 32BLeader
Neither model has a result on 5 boards here

What each workload costs

Three fixed token mixes at each model's listed price, so per-token rates become a bill.

Chat turn

1K input + 500 output tokens

Qwen2.5 72B Instruct$0.0006Fits in one request
Qwen3 32B$0.0003Fits in one request

Qwen3 32B costs less

Repository review

50K input + 3K output tokens

Qwen2.5 72B Instruct$0.019Exceeds the context window
Qwen3 32B$0.0082Fits in one request

Qwen3 32B costs less

Agent loop

200K input + 10K output tokens

Qwen2.5 72B Instruct$0.076Exceeds the context window
Qwen3 32B$0.032Exceeds the context window

Qwen3 32B costs less

Listed OpenRouter rates, without cache or batch discounts. A workload that does not fit a model's context window would need splitting.

Specification differences

The leading value in a row is marked; a row with a missing value names no leader. Our two rows are marked in red.

Specifications of Qwen2.5 72B Instruct and Qwen3 32B
SpecificationQwen2.5 72B InstructQwen3 32B
MakerQwenQwen
WeightsOpen weightsOpen weights
ReleasedSeptember 19, 2024April 28, 2025
Context window33K131K
Input $/Mtok$0.36$0.14
Output $/Mtok$0.40$0.40
Coding indexNot benchmarkedNot benchmarked
Independent measurements behind it00
Cost per solved taskNot computableNot computable
SourceOpenRouter listingOpenRouter listing

Questions

Is Qwen2.5 72B Instruct better than Qwen3 32B for coding?

Neither model has a coding score yet, so the two cannot be ranked on coding. Pick Qwen2.5 72B Instruct for work measured by LiveCodeBench (Artificial Analysis run) or LMArena Text: coding, where it leads.

Which is cheaper, Qwen2.5 72B Instruct or Qwen3 32B?

Qwen2.5 72B Instruct costs $0.36 per million input tokens and $0.40 per million output tokens; Qwen3 32B costs $0.14 and $0.40. Cost per solved task is known for neither of them.

Which has the bigger context window, Qwen2.5 72B Instruct or Qwen3 32B?

Qwen3 32B, with 131K tokens against 33K.

How many benchmarks is the Qwen2.5 72B Instruct vs Qwen3 32B comparison based on?

38 benchmarks have a result for both models as of October 22, 2026. Qwen2.5 72B Instruct's coding index rests on 0 independent benchmarks and Qwen3 32B's on 0. Where only one model has a result, the page says so instead of naming a winner.

Compare next